Hillshade Tool Returns Strange Grid

[kwatter48]

I am trying to create a hillshade from a mosaic dataset I created from DEMs for portions of Massachusetts, Connecticut, and Rhode Island. My mosaic raster looks correct, but the hillshade tool creates a strange grid pattern not at all representing the elevation data in the mosaic. I am not sure what would cause this or how to resolve it. Thoughts? 

 

Mosaic DEM above

Hillshade product from tool.

[Robert_LeClair]

Interesting!  So a question for you - what is the coordinate system of the raster mosaic?  Is it a geographic or projected coordinate system?  If geographic (i.e. WGS 1984 for example), I would recommend you run the Project Raster GP tool first to UTM or Stateplane.  Then run the Raster Function again.  Does the output change to expected results?

[PLadd]

@kwatter48 did you ever resolve this?  I am running into a similar problem (also in CT).

At a scale of around 1600, it looks normal:

But if I zoom in one more roll of the mouse wheel, it looks horrible (like I'm watching reruns on an old black & white TV using an antenna):

This is how bad it gets at 1:240:

Same area/scale from an older hillshade created using ArcMap and an older DEM:

To answer @Robert_LeClair question, as least for me, I am using CT Stateplane projection.

[Robert_LeClair]

Interesting...with regards to your Mosaic Dataset, have you built overviews for the MDS?  Also, can you share the data with me so I may test on my side to figure out what's going on?  Thx!

[BillFox]

maybe bump this setting way up > "Maximum Number of Rasters Per Mosaic"

or do that and then blow out the existing overview and recreate them

or do that and create a fresh file geodatabase, fresh mosaic dataset, etc.

[PLadd]

The steps I was using to create a Hillshade layer, detailed below, were not the best approach.

• From over 200 smaller DEMs, I had created a city-wide raster.tif DEM.
• Then I used this city-wide DEM to build the Hillshade raster, first using the Hillshade function and then trying the GP Tool Hillshade. 
• Both yielded the same fuzzy result.

A colleague at CT ECO, who had built out the Connecticut statewide hilldshade, recommended creating a Mosaic Raster dataset referencing the individual DEM tiffs.  Then using the GP Tool Hillshade, I was able to create a nice clean Hillshade layer.

The result is much better at 1:240.  Hope this can help somebody else. . . .

This was how the old/wrong approach looked:

Thanks for the replies just the same.